atlas_artifact

Provides a Data Source to access to deployment artifacts managed by Atlas. This can be used to dynamically configure instantiation and provisioning of resources.

Example Usage

An artifact can be created that has metadata representing an AMI in AWS. This AMI can be used to configure an instance. Any changes to this artifact will trigger a change to that instance.

# Read the AMI
data "atlas_artifact" "web" {
    name = "hashicorp/web"
    type = "amazon.image"
    build = "latest"
    metadata {
        arch = "386"
    }
}

# Start our instance with the dynamic ami value
# Remember to include the AWS region as it is part of the full ID
resource "aws_instance" "app" {
    ami = "${data.atlas_artifact.web.metadata_full.region-us-east-1}"
    ...
}

Argument Reference

The following arguments are supported:

• name - (Required) Name of the artifact in Atlas. This is given in slug format like “organization/artifact”.

• type - (Required) The type of artifact to query for.

• build - (Optional) The build number responsible for creating the version of the artifact to filter on. This can be “latest”, to find a matching artifact in the latest build, “any” to find a matching artifact in any build, or a specific number to pin to that build. If build and version are unspecified, version will default to “latest”. Cannot be specified with version. Note: build is only present if Atlas builds the image.

• version - (Optional) The version of the artifact to filter on. This can be “latest”, to match against the latest version, “any” to find a matching artifact in any version, or a specific number to pin to that version. Defaults to “latest” if neither build or version is specified. Cannot be specified with build.

• metadata_keys - (Optional) If given, only an artifact containing the given keys will be returned. This is used to disambiguate when multiple potential artifacts match. An example is “aws” to filter on an AMI.

• metadata - (Optional) If given, only an artifact matching the metadata filters will be returned. This is used to disambiguate when multiple potential artifacts match. An example is “arch” = “386” to filter on architecture.

Attributes Reference

The following attributes are exported:

• id - The ID of the artifact. This could be an AMI ID, GCE Image ID, etc.
• file_url - For artifacts that are binaries, this is a download path.
• metadata_full - Contains the full metadata of the artifact. The keys are sanitized to replace any characters that are invalid in a resource name with a hyphen. For example, the “region.us-east-1” key will become “region-us-east-1”.
• version_real - The matching version of the artifact
• slug - The artifact slug in Atlas

See the source of this document at Terraform.io